ANNEX 2 NON-CONFIDENTIAL VERSION

APPENDIX 1: OFCOM’S INCONSISTENT REFERENCES TO RELEVANT MARKETS

1.1 One of the most surprising features of Ofcom’s approach to market definition in the Consultation Document is the inconsistent (and largely haphazard) way in which market definitions are stated, and the fact that a number of the statements on the market definitions that Ofcom has reached comprise products that are not actually supplied to consumers.

1.2 Set out below is an indication of the extremely large variety in the ways in which Ofcom expresses its findings in relation to relevant markets, separated according to whether they relate to sports, film or basic pay TV-only services.1

Sports • “There is a narrow economic market for premium sports retailing (sic) in the UK, which does not include FTA or basic-tier TV channels”2 • “…the retail market for packages containing premium sports…channels“3 • “It remains likely that there are separate retail and wholesale markets for the supply of premium sport… channels. The market for premium sports channels is likely to include both Sky Sports and Setanta, although we cannot rule out a narrower market for Sky Sports alone”4 • “the premium sports content market”5 • “the sports…market”6 • “the wholesaling of premium sport pay TV channels”7 • “a separate market for the retailing of pay TV packages containing premium sports content”8 • “the premium sports retailing market”9 • “in the retailing of premium sports rights”10 • “the wholesaling of premium sports channels”11

1 We note that the following quotations are Ofcom statements in relation to the relevant markets it considers exists (i.e., it does not include references in the Consultation Document to findings on market definition by other regulators in the past). 2 Paragraph 4.37 of Annex 13. 3 Paragraph 5.54. 4 Paragraph 5.23. 5 Paragraph 5.58. 6 Ibid. 7 Paragraph 4.48. 8 Paragraph 5.15. 9 Paragraph 5.32. 10 Paragraph 5.33. 11 Paragraph 5.39. ANNEX 2 NON-CONFIDENTIAL VERSION

• “the wholesaling of premium sports rights”12 • “the wholesaling of premium sports content”13 • “the premium sports pay TV market”14

Movies • “…the retail market for packages containing premium … movies channels (sic)”15 • “…retail and wholesale markets for the supply of premium… movie channels”16 • “…the premium movies retail market”17 • “the premium movies market”18 • “the… movies market”19 • “… premium movie bundles”20 • “the wholesaling of subscription premium movie channels”21 • “the retail premium movies markets”22 • “the premium movies retailing market”23 • “the wholesale market for premium movie content”24 • “the wholesale market for premium movies”25

Basic pay TV services • “It seems likely that basic-tier pay TV and free-to-air TV are also in separate retail markets”26 • “…the wholesale market for basic-tier pay TV”27

12 Paragraph 5.43. 13 Paragraph 5.52. 14 Paragraph 5.63. 15 Paragraph 5.54. 16 Paragraph 5.23. 17 Paragraph 5.54. 18 Paragraph 5.58. 19 Ibid. 20 Paragraph 4.82. 21 Paragraph 4.90 of Annex 13. 22 Paragraph 5.53. 23 Paragraph 5.63. 24 Paragraph 5.64. 25 Paragraph 5.69. 26 Paragraph 5.23. 27 Paragraph 5.24.

2 ANNEX 2 NON-CONFIDENTIAL VERSION

• “packages containing only basic-tier TV channels”28 • “stand-alone basic-tier pay TV”29 • “wholesaling of basic TV channels”30 • “subscriptions to TV packages containing only basic channels”31 • “stand-alone basic pay TV services”32

Other • “the market for premium wholesale channels.”33 • “the wholesale channel provision market”34

1.3 In relation to a number of these statements, we note that these comprise services that are not actually supplied to consumers. Most notably, as Ofcom is aware35 Sky’s premium sports channels and premium film channels are provided to consumers as part of broader packages that also include basic pay TV channels. It is, therefore, extremely odd (to say the least) to reach a market definition that is:

“a narrow economic market for premium sports retailing (sic) in the UK, which does not include FTA or basic-tier TV channels”. 36

1.4 It is similarly strange to conclude that there is a relevant retail market for either “premium movies” or “premium movie channels”.

1.5 Furthermore, a significant problem with Ofcom’s approach is that it does not address pay TV packages that include both premium sports and premium film channels (both at the wholesale levels). All the Ofcom’s various approaches to market definition refer to one or other of these types of channel.

28 Paragraph 5.54. 29 Ibid. 30 Paragraph 5.56. 31 Paragraph 5.71. 32 Paragraph 5.88. 33 Paragraph 6.77. 34 Heading preceding Paragraph 6.22. 35 See, for example, Paragraph 4.31. 36 Paragraph 4.37 of Annex 13.

3 ANNEX 2 NON-CONFIDENTIAL VERSION

APPENDIX 2: FAILURE TO EXPLAIN THE CONCEPT OF “MUST HAVE” CONTENT AND CHANNELS

2.1 A significant pillar in Ofcom’s approach to the assessment of relevant markets and market power is the proposition that some types of programmes, services and television channels are “must have”. For example, Ofcom states:

“We have therefore considered whether narrow economic markets exist for specific ‘must have’ content (premium sport and movies) at both the retail and wholesale level.”37

2.2 This statement makes it clear that one of the principal starting points of Ofcom’s approach to market definition in this case is a presumption that there is such a thing as ‘must have content’. In such a situation, it might be expected that Ofcom would explain what it means when it uses this term and it is therefore highly surprising that Ofcom provides no such explanation.

2.3 In relation to “the retail level” it is wholly unclear what it might mean for television programmes to be “must have”. In the absence of any discussion or definition of the concept by Ofcom, it is possible only to speculate on what might be intended by it.38

2.4 As a starting point, it is certainly possible to say what it does not mean: clearly, the concept of “must have” cannot be intended to mean ‘essential to survival’, and it is improbable that Ofcom means that consumers would be willing to pay infinite prices to watch particular TV programmes.

2.5 An attempt to discern the meaning of this term from the context in which it generally is used in Ofcom’s Consultation Document leads to the conclusion that there are potentially two (mutually exclusive) meanings that Ofcom attaches to it. Either:

(i) that some particular types of content are valued very highly by all (or perhaps nearly all) consumers. Such an interpretation is suggested, for example, by the total lack of qualification around the use of the term in relation to consumers; or

(ii) that at least some consumers have a high willingness to pay to watch particular types of programmes on TV. Such an interpretation is suggested by significant parts of Ofcom’s consumer research, which asks consumers essentially how much they value watching particular

37 Paragraph 1.3 of Annex 13. 38 We overlook in this appendix entirely inexplicable uses of the term by Ofcom – for example the observation that, according to Ofcom, around 88% of consumers in an Ofcom survey regarded “content” as being “must have” or, even more bizarrely, that around 53% of pay TV subscribers surveyed regarded “price/cost” as “must have”. (See Figure 12 of the Consultation Document.) ANNEX 2 NON-CONFIDENTIAL VERSION

types of programming on a scale of “must have” through to “nice to have” - in other words, seeking to determine a preference ranking.

2.6 Clearly, the former interpretation is entirely implausible and belied both by common sense, and Ofcom’s consumer research. Even amongst consumers who said that specific programmes or types of programme were important to them, the programme most often cited as “important” was Coronation Street (38% of respondents), and the most mentioned genre of programme was sports with 45%, citing it as “important”, ahead of soap operas with 39%. The proportion citing each of these as “must have” is not available from the information provided to Sky, but it is likely to be significantly lower.39

2.7 If the latter interpretation, then, is what Ofcom means when it refers to content being “must have”, the proposition is both (a) trite and (b) irrelevant to market definition. It implies simply that there is a demand curve for particular types of programming, which is entirely what common sense would suggest. The fact that there are dedicated fans of particular types of programming, who have a high willingness to pay to view particular types of content on TV, is entirely unremarkable. The same can be said of most goods, and most TV programmes – though, as far as Sky is aware, the epithet of “must have” is not generally applied to other types of goods or TV programmes simply because some people like them a lot.40

2.8 Moreover, as explained in Annex 2, focusing on the preferences of those consumers for whom watching particular types of content on TV is very important is an erroneous starting point in an analysis of market definition. It is the preferences of those least attached to products (marginal consumers) rather than the preferences of those who are most attached (infra-marginal consumers, or dedicated fans) that matter for market definition. Accordingly, if this interpretation of what Ofcom means when it presumes that some content is “must have” and uses this presumption as one of the principal starting points for its analysis of market definition, then it leads directly to Ofcom’s perpetration of the ‘toothless fallacy’.

39 In relation to the two types of programming that appears on pay TV channels that Ofcom appears to regard as falling into the “must have” category (namely live FAPL football coverage, and movies produced by the six major Hollywood studios in the pay TV window) all the available evidence points to the fact that the number of consumers with high willingness to pay to watch this programming on TV is relatively limited. This should not be surprising. In spite of the large amount of media coverage that it receives, following FAPL football is a minority interest among people living in the UK – Ofcom’s own analysis, for example, shows that it appeals to only a quarter of the UK population (among which not all will have a high willingness to pay). (See Ofcom and Human Capital, “Premier League Football: research into viewing trends, stadium attendances, fans’ preferences and behaviour, and the commercial market”, Figure 21). While movies are of broader appeal among the population it seems likely that the large range of alternatives for viewing them diminishes significantly the willingness to pay to watch them on TV in the period between 12-27 months after cinematic release (i.e., in the pay TV window). 40 This is the case even if it is found that there is a significant number of consumers who have a high willingness to pay for a particular product.

2 ANNEX 2 NON-CONFIDENTIAL VERSION

2.9 It is equally unclear what “must have” might mean at the wholesale level and again Ofcom provides no explanation, in spite of asserting that “Sky Sports therefore continues to be a “must have” product”.41 Again, it is possible to rule out certain interpretations of the term as being absurd – for example it cannot mean that retailers would go out of business if they were unable to retail particular television channels (so it cannot mean ‘essential for survival’42) nor can it mean that retailers would be prepared to pay infinite amounts to carry such channels in their pay TV packages. Clearly, both propositions are nonsensical.

2.10 In this case, the proposition would seem to be closely linked to the “key driver” theory, which is that such channels act to attract those consumers who are interested in the content carried on the relevant channels to subscribe to pay TV services. Again, however, such a proposition is both entirely unremarkable and contributes little to an assessment of relevant markets at the wholesale level.

2.11 If these interpretations are not what Ofcom means when it uses this term, then respondents are left in the invidious position of being unable to address a concept which clearly plays a central role in Ofcom’s thinking on the matters addressed in the Consultation Document.

Conclusion

2.12 It is clear that the concept of “must have” content plays a key role in Ofcom’s thinking in the Consultation Document, particularly with regard to the issues of market definition and market power. The meaning of this term is not self- evident and Ofcom does not explain what it means by it. To the extent that it is possible to discern what Ofcom means, it does not appear to have the relevance to market definition that Ofcom ascribes to it. Ofcom should clarify what it means by “must have” and the relevance of this concept to market definition in this case.

41 Paragraph 5.46 of Annex 13. Ofcom appears less certain about its proposition elsewhere in Annex 13. For example at paragraph 5.49 Ofcom states that “Sky Sports is almost certain to be a ‘must have’ product for Virgin Media”, and, at paragraph 5.47, Ofcom states that “Sky Sports is likely to be a ‘must have’ product”. 42 It is notable in this context that around [CONFIDENTIAL]% of Virgin Media customers who subscribe to pay TV services do not subscribe to Sky’s sports and film channels.

3 ANNEX 2 NON-CONFIDENTIAL VERSION

APPENDIX 3: FAILURE TO EXPLAIN WHAT OFCOM MEANS BY “PREMIUM CONTENT”

3.1 The concept of “premium content” is central both to Ofcom’s Consultation Document in general, and issues such as market definition and market power in particular. (The term ‘premium’ occurs over 300 times in the Consultation Document itself and nearly 250 times in Annex 13.) It is therefore surprising that Ofcom provides no explanation of what it means when it uses this term. Ofcom appears to treat the term as though its meaning is self-evident. This is not the case.

3.2 In relation to “premium sports content” it is evident that Ofcom regards live broadcast of FAPL football matches as falling into this category. It is, however, impossible to tell from the Consultation Document whether Ofcom regards any other programming as comprising “premium sports content”.43

3.3 In relation to “premium film content”, or “premium movies”, the only clue at what Ofcom means in relation to this term is provided in a note to Figure 27, which states:

“we have taken premium movies to mean first-run films”

3.4 Given that the term “first-run” simply relates to the first method of distribution of a film – which might be in cinemas, via DVD, on a VoD/PPV service, on a pay TV channel or on a free to air channel this definition provides no meaningful explanation of what Ofcom means when it uses the term “premium movies” in relation to pay TV services.

3.5 There are some indications that Ofcom may have fallen into the trap of believing that the nature of content is determined by the type of channel on which it is carried – i.e., that programming is “premium” if it is broadcast on a television channel defined as a premium (pay TV) channel, and “basic” if it is broadcast on a channel defined as a basic pay TV channel or a free to air television channel.44

3.6 If this is what Ofcom has in mind, then it would involve a bizarre definition of ‘premium’ and ‘basic’ content. For example, it would mean that Ofcom regarded Aerobics Oz Style (which is broadcast on Sky Sports) as ‘premium content’, but not the final of the football World Cup (which is broadcast free to air). Equally, such a definition would mean that programming such as Planet Earth, Jane Eyre, or Shameless would not be regarded as being ‘premium content’. Such an approach would, therefore, be nonsensical.

43 While the heading preceding Paragraph 3.65 is “premium sport”, the discussion that follows is about (a) FAPL football and (b) sports in general, and therefore not informative about what Ofcom means when it uses this term. 44 See, for example, Ofcom’s discussion of basic content and channels under the heading of “General news and entertainment” (sic), at Paragraphs 3.72-3.73. ANNEX 2 NON-CONFIDENTIAL VERSION

Key content

3.7 In a similar vein, Ofcom uses the terms “key content” and “key content rights” throughout its Consultation Document as though the meaning of these terms are self-evident. They are not. Beyond the inference that Ofcom considers the broadcast of live FAPL football matches, and broadcast of films from the six major Hollywood movie studios in the pay TV window, to be “key content” Ofcom provides no indication at all of what else it considers might fall into such a category. Similarly, it provides no indication at all as to what it considers to be “key content rights” beyond rights associated with such programming.

3.8 Such a lack of transparency in relation to concepts that clearly play a central role in Ofcom’s thinking on these matters is particularly unsatisfactory.

Premium channels

3.9 We note here that Ofcom also fails properly to define what it means when using the term “premium channels”. In Sky’s view, this failure is somewhat less significant than the failure in relation to “premium content”, as the concept of premium (and basic) pay TV channels is somewhat more commonly understood, although it remains far from clear-cut. For example, it is not clear whether a television channel (such as Setanta Sports) that is available in a basic-tier package from one retailer (Virgin Media) and as a premium service from another retailer (Setanta via DTH) is a basic or a premium pay TV channel.

3.10 Ofcom’s description of premium sports channels, however, is particularly unclear. Ofcom attempts to set out what it means by “premium sports channels” at Paragraphs 5.25 and 5.26 of the Consultation Document. These paragraphs are as follows:

“A ‘premium sports’ pay TV service is usually defined as one that provides live access, often on an exclusive basis, to a specific set of highly-valued sports events. For the purposes of this work, we have considered packages of premium sports that include access to live FAPL matches.” 45

3.11 Sky considers that this explanation is inchoate:

• it is not clear where “a ‘premium sports’ pay TV service” is “usually defined” as taking the definition that Ofcom cites. Ofcom provides no references for this assertion;

• Ofcom does not explain why it refers to events “often” being broadcast on an exclusive basis when exclusive broadcast of programming (both sports and non-sports) is the norm in television broadcasting;

45 Later in the Consultation Document, Ofcom also adds: “The other key distinction between free to air sports and premium sports (sic) relates to the quantity of programming available” (Paragraph 5.30). 2 ANNEX 2 NON-CONFIDENTIAL VERSION

• Ofcom does not cite any examples beyond “live FAPL matches” as events that it considers to comprise the “specific set of highly-valued sports events” to which it refers; and

• the meaning of the phrase “packages of premium sports” is unclear.

3.12 Moreover, Ofcom’s definition would appear inadequate. For example:

• both ITV1 and BBC1 “provide live access, often on an exclusive basis, to a specific set of highly-valued sports events”. Clearly these are not “premium sports channels”; and

• Setanta Sports provides “access to live FAPL matches”; however, as noted above, this is carried as a basic channel in Virgin Media’s pay TV packages.

3.13 Accordingly, Ofcom fails properly to define premium sports channels in its Consultation Document.

3.14 Ofcom’s definition of “premium movie channels” is clearer than that for “premium sports channels”. Ofcom defines a “premium movie channel”46 by reference to what it describes as its “primary characteristic” which is taken to be that:

“… it provides access on a subscription basis to first-run movies47 from the six major Hollywood studios.”48

3.15 This simply defines a “premium movie channel” as a television channel which carries films from (presumably one or more of) the six major Hollywood studios in the pay TV window. (By definition such a channel will provide access to such films on a subscription basis.)

Conclusion

3.16 It is evident that the term “premium content” plays a key role in Ofcom’s thinking in relation to the issues raised in the Consultation Document – particularly with respect to the issues of market definition and market power. As in the case of “must have” content, the meaning of this term is not self- evident and Ofcom does not explain what it means when it uses it. Such a lack of transparency is a material flaw in Ofcom’s Consultation Document.

46 For reasons that are unclear, Ofcom refers instead to a “premium movies pay TV service” in the actual body of its definition. 47 Again, the reference to “first-run movies” is erroneous. See Paragraph 3.4 above. 48 Paragraph 5.36 of Annex 13 to the Consultation Document. 3 ANNEX 2 NON-CONFIDENTIAL VERSION

APPENDIX 4: TREATING PAST FINDINGS BY COMPETITION AUTHORITIES AS “PRECEDENT”

4.1 Ofcom places significant weight on past findings in relation to competition inquiries in cases involving pay TV both in the UK and other countries. These are described as “precedents” for the current case in the Consultation Document, and form the starting point for Ofcom’s analysis.49 Effectively, Ofcom’s approach is to seek to examine whether such past findings continue to hold.

4.2 At the outset, it is important to recognise that previous findings do not comprise “precedent”, a term which has a specific legal meaning and refers to binding legal decisions. Previous findings as to relevant markets are not binding on future cases, but derive from analysis of facts of particular cases at a given point in time. 50

4.3 Moreover, previous findings are only informative in respect of a subsequent case to the extent that (a) they were well-founded at the time that they were reached, and (b) if so, that the underlying considerations are the same or generally similar to a subsequent case. Ofcom does not appear to have attempted to determine these matters in relation to the past cases that it cites.

4.4 There are clear a priori grounds for considering that many of the past findings are unlikely to be informative in relation to the current case.51 For example:

(i) a number of the findings are from other countries in which the market circumstances are significantly different from those that are found currently (and prospectively) in the UK – for example, no other country in Europe has a tax-funded broadcaster which operates at the scale of the BBC;

(ii) by common consent – including a large number of Ofcom publications - the television sector in the UK is subject to rapid change. Accordingly, it would not be expected that findings in relation to past cases that were

49 Somewhat bizarrely, Ofcom appears to regard findings reached as long ago as 1996 and 1999 as “recent” (as indicated by the section heading preceding Paragraph 2.4 of Annex 13). Clearly, this is not the case. As set out in Sky’s Response to the Complaint, the significant changes in the competitive landscape that have occurred since 2002 mean that it is unsafe even to rely on the OFT’s 2002 decision as “precedent” in the current case. By contrast, the Competition Commission’s BSkyB/ITV report is so recent that the final report was published after Ofcom’s Consultation Document. 50 The Court of First Instance, in Joined Cases T-125/97 and T-127-97, Coca-Cola v Commission 22 March 2000, established that the market must always be defined in any particular case by reference to the facts prevailing at the time and not by reference to “precedents” (Paragraph 82). 51 See Paragraphs 3.5 and 3.6 of Sky’s Response to the Complaint. Ofcom appears to have failed to have regard to Sky’s submission on this matter. ANNEX 2 NON-CONFIDENTIAL VERSION

more than a few years old would be informative in relation to current cases in this sector; and

(iii) many past findings have been reached in circumstances where there was either no ability or no incentive to challenge those findings. Notable in this respect are findings in foreign merger cases in which conditional approval to merge has been given, and the OFT’s 2002 decision, which found no contravention of Chapter II of the Competition Act, which is discussed further below.

4.5 A further reason for taking a cautious approach to whether past findings are relevant to the current case is that a significant number of them were made on the basis of analysis and evidence that do not accord with either current guidance on market definition,52 or the well-established requirement53 that market definition must consider the facts in any given case and the whole economic context of the case.54 It is inappropriate to cite the conclusion of an inquiry without any regard at all to how that conclusion was reached.

4.6 Sky notes that Ofcom makes no reference to findings which would tend to cast doubt on the “preliminary conclusions” reached in the Consultation Document in relation to market definition. For example, there is no reference to the judgment of the Restrictive Practices Court case in the 1999 Premier League case, nor the Competition Commission’s 2003 Carlton/Granada merger report. The latter, for example, stated:

“The previous [Competition Commission report on the merger situation between United News and Media, Carlton and Granada] also saw pay-TV as a segment in the overall market for television audiences (see paragraph 2.48)…. [It also] recorded an increasing convergence between free-to-air and pay-TV services (see paragraph 4.27). That convergence has continued over the past three years, and the launch of Freeview has, by introducing non-terrestrial channels which are free-to-air, further blurred the distinction between pay-TV and terrestrial channels, to the extent that they can be seen as differentiated products within the same market.” 55

52 Notable in this respect are past findings which rely almost entirely on the identification of differences in product characteristics between pay TV and free to air television, which is an insufficient basis for determining whether two products are supplied in separate relevant markets. 53 See, for example, the Judgment of the Competition Commission Appeal Tribunal in Case No. 1005/1/1/01, Aberdeen Journals Limited v. Director General of Fair Trading (Aberdeen Journals), 19 March 2002, at Paragraph 97. 54 In particular, a significant number of European Commission decisions simply recite a standard formulation in order to justify a conclusion that pay TV and free to air services comprise separate relevant markets. Such an approach does not accord with the requirement to have regard to the particular facts of a case. 55 Paragraph 5.33.

2 ANNEX 2 NON-CONFIDENTIAL VERSION

4.7 Furthermore, the European Commission’s analysis of the impact of free to air services on pay TV providers in Germany in the context of market definition is of particular note. The Commission stated:

“The German pay-TV market is difficult to develop because there is a strong market for free TV in Germany. Pay-TV and free TV are separate markets, as has already been explained. There is, nevertheless, some interaction between them. The more varied and attractive the programmes offered by the free broadcasters, the less incentive there is for viewers to subscribe to pay-TV as well.”56

4.8 If Ofcom is to cite past findings as being relevant to the current case, in order to be balanced and fair-minded it should (a) report accurately the nature and basis for those findings; (b) ensure that the underlying facts of each case are relevant to the current case; (c) ensure that the findings were reached on the basis of analysis and evidence of a standard that would be regarded as acceptable currently; and (d) ensure that the market definitions reached were established as a result of examination of the facts in any given case, the whole economic context of the case, and the particular circumstances of that case.

4.9 In Sky’s view, the presumption should be that findings reached more than ten years ago – such as the OFT’s 1996 review, which deals with a sector which had only just been established in the UK – have no relevance in current circumstances and should be entirely disregarded, and cases from other countries given little or no weight due to the dissimilarities in circumstances as between countries.

4.10 Finally, in light of the emphasis that Ofcom places on the consistency of its “preliminary conclusions” on market definition with those reached by regulators in the past, it is notable that as far as Sky is aware no UK regulator has ever defined basic-only pay TV packages to comprise a separate relevant product market, as Ofcom does in the Consultation Document.

Ofcom’s reliance on the OFT’s 2002 decision

4.11 The extent to which Ofcom has treated the conclusions in relation to market definition and market power which are set out in the OFT’s 2002 decision as a reliable starting point for its own assessment is evidenced by Ed Richards’ public statement that:

“If you look at [the OFT’s] conclusions to a very substantial piece of work, they say, in their conclusions, that Sky is dominant in the provision of wholesale premium content, and that is the base from which I am working.

56 Paragraph 56, Case IV/M.0037 BSkyB/KirchPayTV (2000), Official Journal C 110. It is evident from the Commission’s analysis in this case that it was not willing for such obvious evidence of the inter-relationship between the two types of services to overturn its intrinsic belief that the two types of services are provided in separate relevant markets because they have different characteristics.

3 ANNEX 2 NON-CONFIDENTIAL VERSION

It may be that, in the work that we do, we conclude that that is no longer the case, but that is the last serious competition authority assessment of that position and, at the moment, I do not see any reason to change that view. In the course of the work we do, we may do, but that is my current view. I am always happy to receive evidence to refute that from any party, but that is the basis of the current view.”57

4.12 Sky considers that the proposition that the OFT’s conclusions in the 2002 Decision constitute a reliable starting point for Ofcom’s analysis in this case is erroneous for the following reasons:

• the conditions of competition in the sector have changed materially since the OFT conducted its analysis. It is important to bear in mind that, while the decision itself was published in December 2002, the analysis conducted by the OFT was backward-looking and covered a period from June 1998 to December 2001.58 Accordingly, the period examined by the OFT is now over six years ago which, given the pace of change in the sector since then, is a considerable period of time;

• the quality of evidence and analysis of market definition and market power in the decision is poor. In particular, the OFT makes several of the same analytical errors that Ofcom has made in its own analysis – for example, (a) a focus on the preferences of infra-marginal subscribers, (b) a failure to consider the trade-offs that consumers make between different television services; and (c) reliance on assertion and conclusions about the “sufficiency” of competitive constraints that were not based on objective benchmarks;

• the 2002 decision found that Sky had not infringed the Competition Act. Accordingly, there was no reason for Sky to challenge it before the CAT. Had the OFT found that Sky had infringed the Competition Act, it is clear that the issue of whether the OFT’s analysis sufficiently supported its conclusion that Sky at that time held a dominant position in a number of narrowly defined relevant markets would have formed a key part of Sky’s challenge to the decision; and

• a careful examination of the report indicates that conclusions in relation to the retail level were ‘tacked on’ to the analysis. The focus of the analysis was entirely on Sky’s market position as a wholesaler of pay TV channels; there is very limited analysis of issues relevant to market definition and market power at the retail level of the sector.

57 Comments made at the first media analysts’ briefing on 19 September 2007. 58 The OFT issued its Rule 14 Notice in December 2001; its analysis did not change materially after that point.

4 ANNEX 2 NON-CONFIDENTIAL VERSION

The Competition Commission’s findings in the BSkyB/ITV case

4.13 In light of the foregoing, it is remarkable that Ofcom effectively ignores or attempts to downplay the findings of the Competition Commission’s report on Sky’s purchase of a 17.9% stake in ITV. The Competition Commission’s report comprised a very significant assessment of the issues relevant to market definition in the current case, and took evidence from a wide range of operators in the industry. Sky was required to provide the Competition Commission with large amounts of relevant evidence, and we expect that ITV was also required to submit a great deal of evidence. ITV argued strenuously throughout the inquiry that free to air and pay TV services compete strongly with each other and the entirety of the Competition Commission’s finding in relation to a significant lessening of competition rests on that proposition.

4.14 Moreover, as noted in Annex 2 (Market definition and market power), the Competition Commission’s inquiry is: (a) a UK case; (b) recent (it is the only full inquiry in the UK in relation to these issues to have occurred since the emergence of Freeview, for example); and (c) to a significant extent forward- looking (given that it was a merger case).

Conclusion

4.15 Ofcom’s approach of relying heavily (and selectively) on the conclusions of past inquiries as a basis for its own approach to the assessment of market definition and market power in this case is entirely inappropriate. Most of those past findings were themselves either significantly flawed, or have been rendered out-of-date by changes that have occurred in the sector since they were made. By contrast, the Competition Commission’s inquiry in the BSkyB/ITV case is recent – it was published only at the end of last year – and relates directly to the sector that Ofcom is itself examining. Accordingly, the past findings that should be accorded greatest weight in Ofcom’s inquiry are those reached by the Competition Commission in the BSkyB/ITV case.

5 ANNEX 2 NON-CONFIDENTIAL VERSION

APPENDIX 5: OFCOM’S ANALYSIS OF PAST SUBSCRIPTION CHARGES, SUBSCRIBER NUMBERS AND QUALITY OF SERVICES IS FUNDAMENTALLY FLAWED

5.1 In relation to each of the candidate product markets, Ofcom undertakes an assessment of the past history of Sky’s subscription charges, Sky’s subscriber numbers and quality of Sky pay TV packages for the period from January 2000 to October 2007. 59 The results of Ofcom’s assessment are contained in Annex 13 of the Consultation Document.

5.2 There are differences in approach between Ofcom’s analysis in relation to (i) Sky’s packages that include Sky’s sports channels; (ii) Sky’s packages that include Sky’s film channels; and (iii) Sky’s basic-only packages and, accordingly, we address them separately in sections B to D below. (As in most of Ofcom’s analysis of market definition and market power, it is entirely unclear how packages that include both Sky’s sports channels and its film channels fit within this approach.) Before doing so, however, we note a number of problems common to all three analyses.

A. General flaws in Ofcom’s analysis

(i) Ofcom’s approach makes an incorrect assumption of demand being stable over time

5.3 The fundamental flaw in Ofcom’s analysis of this issue is that its approach in each case is based on an erroneous assumption that demand for Sky’s pay TV services (all else being equal) was stable throughout the period examined. In essence, Ofcom’s argument is that (absent quality changes) an increase in prices should be observed to result in a reduction in demand for pay TV services as a result of movement along a stable demand curve. Even if changes in quality are ignored, the fact that Ofcom observes both increases in prices and increases in subscriber numbers during this period should have indicated to Ofcom the immediate problem with such a belief; it is incompatible with an assumption that the demand for pay TV services has been stable over this period.

5.4 Such an observation should, of course, not be surprising. The period covered was one of considerable change – it starts just over a year after the introduction of digital television services in the UK, and only seven months after the introduction of free digital set top boxes by Sky. It also includes the introduction and promotion of PVRs by Sky and cable companies, and the introduction of HDTV services. Accordingly, it covers a period in which the defining characteristic has been the introduction of new services and their take-up by consumers, in which – as a result of factors such as growing awareness of services, increasing marketing efforts, and changes in the

59 Sky provided data on its subscription charges and subscriber numbers in response to Question 1 of Ofcom’s information request dated 22 October 2007. ANNEX 2 NON-CONFIDENTIAL VERSION

consumer proposition – demand for Sky’s services has been constantly shifting.60

5.5 Sky continually sets itself ambitious targets to increase its subscriber base and during the period examined by Ofcom has always met or exceeded those targets. As a result, the expected effect of either (a) increased value-adjusted subscription charges (in real terms) – if in fact such increases may be observed;61 or (b) increased competition from substitute services, would not be absolute falls in subscriber numbers, but reductions in the growth of Sky’s subscriber base below what would have been achieved in the absence of such changes.

5.6 Moreover, more recently, by common acknowledgment – including from Ofcom – there have been significant shifts in consumers’ tastes, which are ongoing.62 Relevant in this respect is the fact that consumers have increasing demands for services which provide them with control, flexibility and mobility with respect to the consumption of audiovisual services.

5.7 It is only possible to draw sensible conclusions about substitutability between services from an analysis of changes in prices and demand if it is possible to separate changes in demand caused by shifts along a demand curve from changes in demand caused by shifts in the demand curve. Ofcom fails to do this, or even to recognise this as an issue.

The relevant benchmark

5.8 In light of this, it is evident that the relevant benchmark against which trends in subscriber numbers must be judged is one of increasing subscriber numbers over time. It is equally evident, however, that examining such issues is a complex matter. In particular, establishment of the counterfactual (what the growth in Sky’s subscriber base would have been in the absence of particular changes) is complex.

(ii) Ofcom’s analyses of trends in charges, quality and subscriber numbers are too cursory to enable sound conclusions to be reached

5.9 The types of issues that Ofcom is seeking to address would normally need to be resolved using an econometric model, having first compiled a dataset of relevant variables. The purpose of such a model would be to isolate the effect on the ‘dependent variable’ (for example, subscriber growth over time) of changes in real, value-adjusted prices from the wide range of other factors

60 It also covers a period in which a rival pay TV retailer (ITV Digital) exited the market, which resulted in many former ITV Digital subscribers switching to Sky. 61 Sky considers that such increases are not observed. 62 See, for example, Paragraph 3.7(f) of Part B of Sky’s Response to the Complaint and Ofcom’s Public Service Broadcasting review.

2 ANNEX 2 NON-CONFIDENTIAL VERSION

that are also likely to affect the dependent variable.63 Clearly, such other factors include:

• other relevant charges for taking up and maintaining a subscription to Sky (all of which have changed over time), which include (among potentially numerous others): o charges for set-top boxes; o charges for installation; o discount offers for new subscribers; o charges for using Sky+; o charges for broadband and telephony services; and o charges for Sky’s multi-room service;

• Sky’s marketing effort and expenditure (in its broad sense – for example, including changes in the packaging and presentation of services to consumers);

• the quality of Sky’s customer service;

• the nature and quality of other services provided to Sky subscribers – for example, Sky+, broadband and telephony;

• general economic conditions;

• specific events, such as the collapse of ITV Digital; and

• the quality, nature and cost of alternatives available to consumers, including other pay TV services, bundled offers by other operators, and, clearly, the aggregate range and quality of free to air television services available.

5.10 Even taking into account the fact that this analysis has been prepared for the purposes of consultation, Ofcom’s approach to the analysis of these issues, however, is risible. It consists of eyeballing charts to determine “trends” in subscriber numbers, looking at changes in subscription charges and making impressionistic statements about both quality changes and changes in the range and nature of potential substitutes available to consumers. Such an approach is wholly inadequate as a method of reaching well-supported, evidence-based conclusions.

5.11 As a result, no evidential weight can be placed on the analysis that Ofcom has conducted in terms of reaching conclusions about the scope of relevant markets.

63 Sky pointed this out to Ofcom in its response to Question 1 of Ofcom’s information request, dated 22 October 2007. It is evident that Ofcom has not had due regard to that response.

3 ANNEX 2 NON-CONFIDENTIAL VERSION

(iii) Inconsistencies between analysis of the three types of service

5.12 The considerable inconsistency between the analyses in relation to each product is also notable. For example:

• the ability to take up Sky+ applies to all subscribers to Sky’s pay TV services, yet is mentioned only in relation to basic-only pay TV subscribers as a quality improvement;

• the availability of market leading broadband services to Sky subscribers is referred to as a quality improvement in relation to subscribers to packages that include premium film channels, but is not referred to in relation to either basic-only subscribers or subscribers to packages that include premium sports channels; and

• Ofcom discusses the potential for subscription being a household level decision (and the possibility of different household members having different preferences) in relation to basic-only pay TV services, but fails to consider such an influence on demand in relation to pay TV services that include premium channels.

(iv) Ofcom’s statements with respect to increases in charges in real terms are misleading

5.13 Ofcom’s conclusions are predicated on a view that charges for such services have been increasing in real terms (i.e., after adjustment for inflation) over time. Ofcom gives the impression that charges have increased continuously in real terms since 2000. For example, Ofcom states:

“we have observed that the price of Sky Sports has increased in real terms in the last six years”64

“Since January 2000, sales of premium movie subscription pay TV packages have increased, and prices have increased in real terms”65 and

“Sky has been able to sustain a growth in the number of residential basic- tier subscribers over the period, despite real increases in headline prices through most of the period”66

5.14 Such statements are over-simplistic and, as a result, highly misleading. As shown in Annex 2, charges for Sky’s most popular pay TV packages have been relatively constant in real terms for several years. To the extent that Ofcom finds real increases in charges, it is evident that these are generated by increases that occurred some time ago.

64 Paragraph 4.36 of Annex 13. 65 Paragraph 4.81 of Annex 13. 66 Paragraph 4.111 of Annex 13

4 ANNEX 2 NON-CONFIDENTIAL VERSION

5.15 Ofcom’s reference to Sky being able to achieve increases in the real price of basic-only packages “through most of the period” is particularly misleading. It is clear that the charge for Sky’s Big Basic package was broadly flat between January 2003 and September 2005, when it rose by a modest amount, and has declined continuously since then. In real terms the charge for this package is lower now than in January 2003.

B. Ofcom’s analysis of packages that include Sky’s sports channels

5.16 In relation to Sky’s packages that include its sports channels, the hypothesis that Ofcom has sought to address is as follows:

“if competition has intensified sufficiently for the market definition to be widened (as compared with previous precedent) then we would expect to see some combination of falling prices, falling subscriber numbers and rising quality.”67

5.17 As a result of its analysis, Ofcom concludes that:

“On balance we believe the evidence presented here suggests that historical precedent with respect to premium sports in the UK (sic) is likely to remain relevant.”68

5.18 The first point of note about this approach is the very significant (and in Sky’s view, excessive) amount of weight Ofcom places on the so-called “starting point” being correct. The hypothesis is that – based on the findings of regulatory inquiries in the past (which appears to mean the OFT’s 2002 Decision) – pay TV packages containing premium pay TV sports channels comprised a separate relevant product market at the start of the period.69 Yet Ofcom appears to have done nothing to assure itself that this finding was correct at the time that it was reached. If it was not, (as Sky considers is the case), then the whole basis for this test is flawed. In other words, if the starting point is wrong, then even if “some combination of falling prices, falling subscriber numbers and rising quality” is not observed, the test would not provide corroborative evidence that there is a separate relevant market for pay TV packages that include Sky’s sports channels today.

5.19 Second, as set out above, Ofcom’s benchmark is wrong. Even if there were no changes in prices and quality of Sky’s services during this period, the expectation would not be that the result of “intensified” competition would be “falling subscriber numbers”, but instead, reduced growth in subscriber

67 Paragraph 4.30 of Annex 13. 68 Paragraph 4.30 of Annex 13. 69 As noted in Appendix 4 (Treating past findings by competition authorities as “precedent”), the OFT’s conclusions in relation to retail markets were essentially ‘tacked-on’ to an analysis that was focused entirely on an analysis of Sky’s position as a supplier of premium pay TV channels, as the key focus of the OFT’s inquiry was investigation of Sky’s wholesale charges for those channels.

5 ANNEX 2 NON-CONFIDENTIAL VERSION

numbers relative to the growth that would have occurred in the counterfactual scenario.

5.20 Third, as set out above, Ofcom’s test has no regard to any of the myriad of other factors that would be expected to affect Sky’s subscriber numbers.

5.21 Finally, Ofcom’s examination of changes in the quality of Sky’s pay TV packages that include its sports channels is both (a) erroneous; and (b) cursory. Ofcom’s assessment of changes in quality of those packages since January 2000 is as follows:

“there does not appear to have been a substantial change to the underlying quality of Sky’s sports offering – although we recognise that this is a more difficult judgment to make. On a simple measure, (sic) although there has been some growth in the overall number of hours broadcast, the number of live FAPL games shown by Sky has actually declined following the 2006 auction.

A more complex question is the impact of changes to the set of basic channels that are sold with Sky Sports, and of the introduction of Sky+. These represent improvements to the quality of the overall bundle and may therefore provide some explanation of Sky’s ability to increase prices and subscriber numbers over the period. However, the changes to the quality of the basic channels is relatively less important to premium sports subscribers than to basic-only subscribers and our consumer research points to premium sports subscribers’ strong interest in the sports component of their subscription. These factors suggest that the changes to the quality of basic channels, while relevant, may not be sufficient to account for the increase in the price of premium sports bundles.”70

5.22 Even taking into account the fact that this analysis was prepared for the purposes of consultation, its meagreness is striking.

5.23 Sky notes the following points in relation to Ofcom’s quality assessment:

• the entirety of Ofcom’s assessment of changes in the quality of Sky’s sports channels since 2000 appears to amount to two observations – one of which, as discussed, below is erroneous. The correct observation is that “there has been some growth in the overall number of hours [of live FAPL matches] broadcast”. In fact, this increased by 256% over this period. On any basis, such a statement cannot form a reasonable basis for Ofcom’s evident view that the quality of Sky’s sports channels is no higher now than it was in January 2000;

• Ofcom’s assertion that “the number of live FAPL games shown by Sky has actually declined following the 2006 auction” is wrong. It has in fact increased from 88 to 92 matches. We find the fact that Ofcom has erred

70 Paragraphs 4.27 to 4.28 of Annex 13.

6 ANNEX 2 NON-CONFIDENTIAL VERSION

in this indicative of the weak evidence base from which it has conducted its analysis;

• it is perhaps not surprising that Ofcom finds it difficult to form a judgment as to how the quality of Sky’s sports channels has changed since 2000 as Ofcom does not appear to have any evidence available to it that would inform this judgement. In particular, Sky has not been asked to provide any relevant evidence. The clear inference is that Ofcom’s “judgment” is based on nothing more than a vague impression as to how Sky’s sports channels have changed, which is a wholly inadequate basis on which to develop properly formed views;

• Ofcom cites no evidence for its view that “changes to the quality of the basic channels is relatively less important to premium sports subscribers than to basic-only subscribers”. This would appear to amount to nothing more than an assertion. Ofcom may be confusing the fact that basic-only subscribers do not value premium sports channels sufficiently to subscribe to them with a belief that households who subscribe to packages that include premium sports channels do not value basic channels. If so, this is a non-sequitur. Equally, the fact that households who subscribe to packages that include premium sports channels value premium sports channels sufficiently to subscribe to a package that includes that type of channel says nothing at all about their preferences in relation to basic channels.

In fact, evidence suggests that the vast majority of subscribers to packages that include premium sports channels do also value other aspects of the services they subscribe to – including the basic channels (and indeed, potentially premium film channels) included in their packages.

5.24 A proper assessment of this issue would have provided Ofcom with evidence in relation to the very significant improvements in the Sky Sports channels since 2000. Such improvements are set out in Appendix 9 (Changes in the quality of Sky’s sports and film channels over time) and include: an additional channel (Sky Sports Xtra), improvements in the range of sports and sports events covered (including the introduction of live coverage of Champions’ League football and all England home cricket Test matches) and a large range of technical innovations that both improve coverage of sports events, and provide subscribers with more ways to watch the programming carried on Sky’s sports channels.

5.25 In addition to improvements to Sky’s sports channels, packages in which those channels are included have been enhanced in ways such as: the inclusion of more basic channels and improvements to the programming carried on basic channels that were already carried in Sky’s packages; the introduction of Sky+ (which enables subscribers to get much more value from their existing subscription); and the introduction of services such as market leading broadband and telephony offers (including free 2MB broadband services for all on-net Sky subscribers).

7 ANNEX 2 NON-CONFIDENTIAL VERSION

C. Ofcom’s analysis of packages that include Sky’s film channels

5.26 In relation to Sky’s pay TV packages that include its premium film channels, Ofcom does not clearly set out the hypothesis that it is seeking to examine – it appears to be a different hypothesis to the case of packages that include premium sports channels. Ofcom’s only explanation is that it is looking at “changes in prices and subscriber numbers for packages containing Sky movie products (sic)” in order to:

”help distinguish whether the high levels of stated switching [found in Ofcom’s application of the SSNIP test] are consistent with a highly competitive market, or one in which market power is currently being exercised.”

5.27 It is entirely unclear how Ofcom considers that the analysis that it has undertaken informs this question.

5.28 Ofcom concludes from its analysis, however, that:

“this tends to suggest that the competition from …potential substitute products is limited.” 71

And that its analysis is:

“suggestive of a narrow market”.72

5.29 As far as can be discerned, Ofcom’s hypothesis appears to be as follows:

Given (a) an increase in availability and falling prices of potential substitutes (e.g., films on DVD and free to air TV channels), (b) increasing charges in real terms for pay TV packages that include Sky’s movie channels; and (c) no “significant change in the quality of the Movies package products (sic) over the period” we should (if the market included such substitutes) observe reductions in the number of subscribers to such packages.

5.30 If this is Ofcom’s hypothesis then, clearly, it suffers from many of the flaws discussed in sections A and B above.

5.31 In particular, it is abundantly clear that there is a very significant issue in relation to pay TV movie services associated with the relevant benchmark. As Ofcom notes, correctly, the number of DTH subscribers to Sky’s movie channels ceased increasing at the end of 2003, and fell during 2007. [CONFIDENTIAL] At the same time, take-up of other types of pay TV services continued to increase. The obvious question, therefore, is whether this decline in subscriber numbers relative to a scenario in which subscriber numbers continued to increase is attributable to a combination of increases in

71 Paragraph 4.81 of Annex 13. 72 Paragraph 5.44.

8 ANNEX 2 NON-CONFIDENTIAL VERSION

real, quality-adjusted charges for such services (if in fact such increases took place), and/or increased availability and falling prices of substitutes for Sky’s movie channels.

5.32 As in the case of Sky’s sports services, the evidence that Ofcom relies on in support of its view that there have, in fact, been increases in real, quality adjusted charges for pay TV packages that include Sky’s movie channels is nugatory as a result of its failure adequately to assess quality changes. In this case, Ofcom’s assessment of the quality of Sky’s movie channels since 2001,73 is as follows:

“the number of films broadcast on Sky Movie channels has remained relatively stable since 2001, although the addition of other services, such as free broadband on Sky (sic) or greater on-demand services on Virgin Media and Sky could be viewed as a quality change.” 74

5.33 The meagre nature of this assessment is striking.

5.34 In fact, a proper assessment of this issue would have provided Ofcom with a significant amount of evidence in relation to the improvements in Sky’s movie channels since 2001. Such improvements are set out at Appendix 9 (Changes in the quality of Sky’s sports and movie channels over time) and include:

• an increase in the percentage of blockbuster films shown on the film channels, from 4% to 6% of films;

• a reduction in the temporal gap between cinematic release and when films are first shown on Sky’s film channels, from 18 months after cinematic release, to 12 months after cinematic release; and

• significant improvements in the presentation of films to subscribers by reorganisation of Sky’s channels into genre-based channels.

5.35 Again, on top of such improvements to Sky’s movie channels, packages in which such channels are included have been enhanced in ways such as the inclusion of more basic channels and improvements to the programming carried on basic channels that were already carried in Sky’s packages, the introduction of Sky+ (which enables subscribers to get much more value from their existing subscription), and the introduction of services such as market leading broadband and telephony offers (including free 2Mb/s broadband services for all on-net Sky subscribers).

5.36 Moreover, as noted above, charges for Sky’s pay TV packages that include “premium” movie channels have not in fact increased in real terms over the

73 It is unclear why Ofcom adopts this different starting point for its analysis of Sky’s movie services. 74 Paragraph 4.72 of Annex 13.

9 ANNEX 2 NON-CONFIDENTIAL VERSION

recent past. Accordingly, it appears probable that in real, value-adjusted terms, charges for such services have actually been declining.

5.37 Given this, the obvious cause of falling subscriber numbers is the wide range, and falling price, of alternatives to Sky’s movie channels, noted by Ofcom. In this sense, it is important to bear in mind a number of issues that suggest that the constraint associated with such alternatives is likely to have strengthened considerably towards the end of the period examined by Ofcom. For example:

• Film Four became a free to air channel in July 2006 – accordingly its impact on the strength of the range of substitutes to Sky’s movie channels will have occurred towards the end of the period examined by Ofcom (the period in which Sky’s movie subscriber numbers have been declining);

• the dramatic increase in the number of films carried on free to air television channels is also more heavily focused on the latter end of the period examined by Ofcom. For example, a significant reason for this increase is the adoption by the commercial terrestrial broadcasters of a strategy of developing a portfolio of sister channels. This strategy change post-dates the emergence of Freeview (which began towards the end of 2002). For example, ITV2 (on which a considerable range of films is broadcast) was converted from a pay TV to a free to air channel only in November 2004. The completion of ITV’s development of its current free to air channel portfolio occurred with the introduction of ITV4 in November 2005;

• a range of additional full time free to air film channels, such as Zone Horror, True Movies, and movies 24 have also emerged only relatively recently. The launch dates of such services on DTH were May 2004, April 2005, and June 2006 respectively;

• ntl and Telewest launched their ‘true VoD’ services in 2005; and

• the decline in the prices for DVDs have been skewed towards the end of the period examined by Ofcom, as shown in Figure 5.1 below.

10 ANNEX 2 NON-CONFIDENTIAL VERSION

Figure 5.1: Home video average price per unit

12

10

8

6 Price per unit £

4

2

0 1997199819992000 200120022003200420052006

* Source: BVA/Sky Analysis

5.38 Of course, without proper analysis, it is not possible to determine conclusively the link between such developments and the falling numbers of subscribers to services that include Sky’s film channels. Nevertheless, common sense suggests that they are unlikely to be wholly unrelated.

D. Ofcom’s analysis of Sky’s basic-only packages

5.39 In this case, Ofcom explains the purpose of its assessment as follows:

“We have examined the changes in prices and subscriber numbers of Sky’s basic-tier pay TV packages to see whether this provides any evidence on how consumers have responded to changes in the relative prices of basic- tier pay TV and free-to-air service, as well as how Sky has adapted its pricing strategy to changes in competitive conditions. As with premium sports and movies, we recognise that Sky is only one of a number of retailers of basic-tier pay TV; however, its subscriber numbers are an important indicator of the way the market has developed.”

5.40 Sky considers that this explanation is essentially unintelligible.

5.41 Moreover, it is impossible to determine what Ofcom concludes from its analysis – particularly in Annex 13 of the Consultation Document. We suspect that Ofcom’s actual view is likely to be, as set out in the statement in the main body of the Consultation Document, that in relation to basic-only pay TV packages:

11 ANNEX 2 NON-CONFIDENTIAL VERSION

“The evidence on consumer’s actual response to historic changes in price is… inconclusive.” 75

5.42 If this is the case, it is clear that no evidential weight can be placed on this analysis in reaching conclusions as to the scope of the relevant market.

5.43 As in the case of the other two types of services, Ofcom’s analysis of changes in the quality of Sky’s basic-only pay TV services is cursory. In this case, the full extent of Ofcom’s assessment is as follows:

“there has been a year on year increase in the number of channels offered to consumers. For example in 2002 the top tier basic package (Family) contained [ _ ] channels and a further [ _ ] time shifted channels. By 2007 the equivalent package (6 mix) contained [ _ ] channels and [ _ ] time shifts. Consumers also now have access to Sky’s and Virgin’s DVR services. ”76

5.44 Accordingly, the entirety of Ofcom’s assessment of changes in the quality of Sky’s basic pay TV services would appear to amount to observations that (a) the number of channels included in packages has increased; and (b) subscribers to basic-only pay TV services now “have access to” PVRs. On any measure, such a cursory assessment cannot form a reasonable basis for conclusions as to changes in quality of Sky’s basic pay TV services since January 2000.

5.45 In particular, it is abundantly clear that an assessment of the quality of television services cannot be based simply on an exercise of counting the number of channels provided. It is not difficult to recognise that one channel with a range of ‘high quality’ programming may be a far better proposition from the point of view of many consumers, than five channels all with ‘low quality’ programming. (Of course, measuring the ‘quality’ of programming itself raises considerable conceptual and practical difficulties.)

Conclusion on Ofcom’s analysis of trends in subscription charges, subscriber numbers and quality of services

5.46 Ofcom’s analysis of trends in subscription charges, subscriber numbers and quality of services in relation to (a) Sky’s packages that include Sky’s sports channels; (b) Sky’s packages that include Sky’s movie channels; and (c) Sky’s basic-only packages, suffers from numerous, significant analytical and evidential flaws. It is entirely clear from the foregoing that its only potential use in relation to assessment of market definition is in highlighting the likelihood that an increasing range of substitutes, and reductions in their prices, is reducing demand for pay TV services that include Sky’s movie channels.

75 Paragraph 5.52 of the Consultation Document. Sky notes the strong implication that the analysis in relation to other services is ‘conclusive’, which is not the case. 76 Paragraph 4.113 of Annex 13. It is difficult to understand why the numbers of channels have been redacted from this paragraph.

12 ANNEX 2 NON-CONFIDENTIAL VERSION

APPENDIX 6: INAPPROPRIATE EMPHASIS ON “KEY DRIVERS” OF SUBSCRIPTION

6.1 A key part of Ofcom’s assessment of product characteristics is a search for what are described as the “key drivers” of consumers’ subscription decisions. For example, Ofcom states:

“We also commissioned some more detailed consumer research which sought to identify the key drivers of demand for premium sports channels.”77

“The key drivers of demand for premium sports pay TV packages are high quality sports content, and live football content in particular, together with the availability of a wide range and choice of attractive live sports content.”78

“The key driver of demand for subscription pay TV premium movie packages appears to be the very wide choice of first run Hollywood films that are always available and accessible to consumers at a zero marginal cost.” 79

6.2 Such an approach, and statements about “the key drivers” or “the key driver” of subscription decisions, are flawed for the following reasons:

(i) consumers have heterogeneous preferences. It is overly simplistic to attempt to determine a small number of reasons – let alone a single reason - why consumers in general subscribe to a particular type of pay TV service. The simple fact is that different people subscribe to particular pay TV services for different reasons;

(ii) this is reinforced by the fact that pay TV subscriptions are household level decisions and most households comprise more than one person. Accordingly, it is likely to be the case that a decision to subscribe to a particular pay TV package reflects a number of different interests on the part of the household’s constituents; and

(iii) Ofcom entirely overlooks the fact that pay TV services provided at the retail level comprise bundles of services (including non-pay TV services, such as broadband, telephony and PVR services). For example, it is entirely evident that a significant number of subscribers to services provided by Virgin Media, Sky, Tiscali and BT are motivated significantly by broadband and telephony services that come bundled with pay TV services.

77 Paragraph 4.16 of Annex 13. 78 Paragraph 4.33 of Annex 13. 79 Paragraph 4.78 of Annex 13. ANNEX 2 NON-CONFIDENTIAL VERSION

In particular Ofcom overlooks the role that basic pay TV channels play in motivating consumers’ subscription decisions. Ofcom’s analysis repeatedly assumes that the only reason that consumers subscribe to pay TV packages that include premium sports and/or premium film channels is to watch programmes on those channels. If consumers were only interested in getting access to premium sports or premium film channels, however, it would be expected that most would subscribe to packages that contain the smallest number of basic channels possible. Yet most Sky customers who subscribe to packages that include premium channels subscribe to basic channels packages that are larger than the smallest basic package available.80

Furthermore, if all that motivated consumers to subscribe to pay TV services were premium sports or premium film channels, it would be entirely irrational for companies like Sky to spend large amounts of money, as Sky has done, on promoting such channels – for example in Sky’s highly successful ‘What Do You Want To Watch?’ campaign, which highlighted the range of content available on channels such as TCM, Sky One, National Geographic and Extreme Sports.

The simple fact is that the value derived by the vast majority of consumers from a pay TV subscription (and other services that may be received along with that subscription) derives from the totality of the set of products consumed, not subsets of that set. It may of course be true that consumers place more weight on some features than others in motivating their subscription decision. But it is a non-sequitur (and entirely erroneous) to consider that those elements on which consumers place more weight are the “key drivers” of their subscription, and that the other elements are of marginal or no significance.81

80 [CONFIDENTIAL] 81 In terms of standard consumer choice theory, consumers will subscribe to a pay TV package if the total utility they gain from consumption of all elements of the package exceeds the monthly subscription charge.

2 ANNEX 2 NON-CONFIDENTIAL VERSION

APPENDIX 7: ERRORS IN OFCOM’S APPLICATION OF THE HYPOTHETICAL MONOPOLIST TEST

A. Ofcom’s application of the SSNIP test to “the price of the sports/film element”

7.1 Ofcom errs in its application of the hypothetical monopolist test by applying the SSNIP test to incremental prices.

7.2 The pay TV services that Sky provides to subscribers are pay TV packages– i.e., collections of channels for which subscribers pay a single monthly charge. Sky’s current pricing grid is as follows:

Basic only Basic plus 1 Basic plus 2 Basic plus 3 Basic plus 4 Sport/Movies Sport/Movies Sport/Movies Sport/Movies pack packs packs packs 1 Mix £16 £26 £34 £38 £41 2 Mix £17 £27 £35 £39 £42 3 Mix £18 £28 £36 £40 £43 4 Mix £19 £28 £36 £40 £43 5 Mix £20 £29 £37 £41 £44 6 Mix £21 £30 £38 £42 £45 * Sky provided this table to Ofcom in response to Question 1 of Ofcom’s information request, dated 22 October 2007.

7.3 A key feature of such an approach to pricing of pay TV services is that ‘components’ within any given package – whether individual channels, or groups of channels – do not have identifiable retail prices. For example, none of Sky One, Sky Sports 1, or the Sky Sports Mix have a monthly subscription charge associated with them.

7.4 A corollary of this is that ‘incremental prices’ – the differences between (for example) the monthly charge for package A, and the monthly charge for package A+B – are not meaningful as prices in their usual sense (e.g., the amount that consumers pay for something). For example, it is not the case that the ‘price’ of two movies packs can be derived by deducting the monthly charge for basic packages from the monthly charge for basic packages plus two movie packs. Subscribers do not pay £17 or £18 for two movie packs; they pay between £34-£38 for a package that includes two movie packs and the number of basic Mixes of their choice.

7.5 That this is the case has been recognised by regulators in the past. For example, Paragraph 427 of the OFT’s 2002 Decision stated:

“The difference between the retail price of the largest basic package and the price of a premium package does not correspond to the revenue that should be allocated to premium channel retail. (sic)”.

7.6 Ofcom’s approach to this issue, however, is highly confused. On the one hand, Ofcom appears to recognise the fact that components within packages do not have retail prices and states: ANNEX 2 NON-CONFIDENTIAL VERSION

“Our approach has been, as far as possible, to apply the HMT test to products that individuals actually consume, thereby avoiding attributing notional prices to individual components of bundles”82

7.7 Yet it then goes on to state:

“for premium pay TV services (which are typically sold bundled with basic pay TV channels)… .to apply the HMT test, we assume the price of the premium package increases, whilst holding the price of basic-tier TV packages constant.”83

7.8 Accordingly, what Ofcom has actually done is to assume that the charge for a “premium pay TV service” can be broken down into two “individual components” (a basic component and a premium component) each of which has a separate notional charge. The notional charge that Ofcom attributes to the ‘basic component’ is its stand-alone charge (i.e., the charge that a subscriber would pay if they chose to subscribe to a basic-only pay TV service); the notional charge that Ofcom attributes to the ‘premium component’ is the ‘incremental price’.

7.9 The correct way of applying the SSNIP test is to evaluate consumers’ reactions to a 5-10% increase in the price of the products and services that are actually available to them. For example, in the case of subscribers to Sky’s ‘Top Tier’ package, the correct issue in the context of market definition is what they would do if the subscription charge for their package rose from its current level (£45 per month) to between £47.25 - £49.50 per month.

7.10 The result of Ofcom’s approach is that far lower price increases were put to consumers in Ofcom’s survey than would be suggested by a correct application of the SSNIP test. Instead of increases of between 5%-10% Ofcom asked consumers how they would react to increases in charges of between 2%-6%.84

7.11 Clearly, however, while erroneous, the result of this approach is to understate significantly the amount of switching that would occur in response to a SSNIP at the full level. As a result, although Ofcom’s approach is far from ideal, in fact it provides good evidence that markets should in fact be defined more broadly than those that Ofcom has defined in its preliminary conclusions.

82 Paragraph 3.14 of Annex 13. 83 Ibid. 84 See paragraphs 4.21 and 4.61 of Annex 13.

2 ANNEX 2 NON-CONFIDENTIAL VERSION

B. Failure to consider whether groups of products comprise a separate relevant market

7.12 Ofcom fails also to follow the standard procedure in market definition of adding products that appear to be close substitutes to the focal product and re-applying the test if SSNIPs are found to be unprofitable. At Paragraph 3.3 of Annex 13, Ofcom recognises that this is how the hypothetical monopolist test is intended to be applied. Ofcom states:

“If [a SSNIP] is unprofitable, then we identify the closest substitute to the focal product, add this product to the focal product and repeat the test.”

7.13 There is, however, no evidence that Ofcom has sought to apply this approach. In particular, there is no evidence that Ofcom applied that test – or sought to apply that test - to any group of products wider than the focal product. For example, Ofcom does not appear to have added Setanta’s package to the focal market (based on the fact that it may be the closest substitute to a package that includes Sky’s sports channels) and repeated the application of the hypothetical monopolist test.

3 ANNEX 2 NON-CONFIDENTIAL VERSION

APPENDIX 8: EXAMPLES OF CHANGES TO SKY’S BASIC PACKAGES IN SEPTEMBER 2005

8.1 In September 2005, Sky replaced its existing basic channel packages with six genre-based “Mixes”, of which subscribers could take any two, four or all six.

8.2 Prior to this change, Sky had offered six basic packages, of which each subscriber would select one: the Family Pack (containing all basic channels retailed by Sky), the Value Pack (containing five channels) and four intermediate packages (each containing fewer than 20 channels).

8.3 The old and new packages are not comparable – the change essentially involved a switch to an entirely new way of providing basic channels to subscribers. Nevertheless, it is instructive as to the very significant improvements in value for money of packages below Big Basic/ Family Pack delivered by this change to compare the subscription charge and channels in packages available pre-September 2005, with examples of what was offered to subscribers at nearby price points after September 2005.

8.4 The tables below give three examples of such comparisons.

1 ANNEX 2 NON-CONFIDENTIAL VERSION

(a) Example 1

Pre-September 2005 Post-September 2005 Post-September 2005

Knowledge Pack, £16.50 per Two Mix example: Knowledge Mix and Style & Culture Mix, Four Mix example: Knowledge Mix, Style & Culture Mix, News & month £15.00 per month Events Mix and Variety Mix, £18.00 per month Channels in previous column plus: Animal Planet Adventure One National Geographic abc1 LIVINGtv2 Animal Planet +1 Animal Planet National Geographic +1 Attheraces Motors TV Bloomberg Animal Planet +1 Sky Travel Bloomberg Paramount CNBC Artsworld (now Sky Arts) Sky Travel +1 Bravo Paramount 2 Discovery Channel Biography Sky Travel Extra Bravo +1 RTE One (in NI) Discovery Channel +1 Discovery Channel Sky Travel Shop British Eurosport RTE Two (in NI) Discovery Realtime Discovery Channel +1 Star Plus British Eurosport2 Sci-Fi Discovery Realtime +1 Discovery Civilisation TV5 Monde Challenge Sky Mix (now Sky Two) History Channel Discovery Home & Health TVEi Challenge +1 Sky One History Channel +1 Discovery Home & Health +1 UKTV Bright Ideas CNBC Sky Sports News Music Choice Discovery Real Time Extra UKTV Documentary E! Sky Vegas Live National Geographic Discovery RealTime UKTV Documentary +1 E4 Star News National Geographic +1 Discovery RealTime +1 UKTV Food E4 +1 TCM Sky One Discovery Science UKTV Food +1 Fox News TG4 (in NI) UK Documentary +1 Discovery Travel & Discovery UKTV History Ftn UKTV Drama UKTV Documentary Wings UKTV History +1 FX UKTV Gold DW-TV UKTV People Hallmark UKTV Gold +1 Fashion TV UKTV People +1 LIVINGtv UKTV Gold 2 History Channel UKTV Style LIVINGtv +1 UKTV Gold 2 +1 History Channel +1 UKTV Style +1 Living UKTV Style Gardens ANNEX 2 NON-CONFIDENTIAL VERSION

(b) Example 2

Pre-September 2005 Post-September 2005 Post-September 2005

Popular Mix, £16.50 per Two Mix example: Knowledge Mix and Kids Mix, £15.00 per Four Mix example: Knowledge Mix, Style & Culture Mix, Kids month month Mix and Variety Mix, £18.00 per month Channels in previous column plus: Bloomberg Adventure One Nick Jr abc1 RTE Two (in NI) Cartoon Network Animal Planet Nickelodeon Artsworld (now Sky Arts) Sci-Fi Cartoon Network +1 Animal Planet +1 Nickelodeon +1 Bravo Sky Mix (now Sky Two) Discovery Channel Biography Nicktoons Bravo +1 Sky One Discovery Channel +1 Boomerang Toonami Challenge Sky Travel Discovery Realtime Cartoon Network Trouble Challenge +1 Sky Travel +1 Discovery Realtime +1 Cartoon Network +11 Trouble +1 Discovery Home & Health Sky Travel Extra Jetix Discovery Channel UKTV Documentary Discovery Home & Health +1 Sky Travel Shop Jetix +1 Discovery Channel +1 UKTV Documentary +1 Discovery Real Time Extra Sky Vegas Live Living Discovery Civilisation UKTV History Discovery RealTime Star Plus Music Choice Discovery Kids UKTV History +1 Discovery RealTime +1 TCM National Geographic Discovery Science UKTV People Discovery Travel & Living TG4 (in NI) National Geographic +1 Discovery Wings2 UKTV People +1 DW-TV TV5 Monde Nickelodeon History Channel E! TVEi Nickelodeon +1 History Channel +1 E4 UKTV Bright Ideas Paramount Jetix E4 +1 UKTV Drama Paramount 2 Jetix +1 Fashion TV UKTV Food Sky One National Geographic Ftn UKTV Food +1 UKTV Gold National Geographic +1 FX UKTV Gold Hallmark UKTV Gold +1 LIVINGtv UKTV Gold 2 LIVINGtv +1 UKTV Gold 2 +1 LIVINGtv2 UKTV Style Paramount UKTV Style +1 Paramount 2 UKTV Style Gardens RTE One (in NI) 3 ANNEX 2 NON-CONFIDENTIAL VERSION

(c) Example 3.

Pre-September 2005 Post-September 2005

Value Pack, £13.50 per month Two Mix example: Variety Mix and Music Mix, £15.00 per month Bloomberg abc1 Paramount 2 Discovery Realtime Bravo Q TV Discovery Realtime +1 Bravo +1 RTE One (in NI) Music Choice Challenge RTE Two (in NI) Sky One Challenge +1 Sci-Fi E! Scuzz E4 Sky Mix E4 +1 Sky One Flaunt Sky Vegas Live Ftn Smash Hits FX TCM Hallmark TG4 (in NI) Kerrang The Amp KISS The Box LIVINGtv The Hits LIVINGtv +1 TMF LIVINGtv2 UKTV Drama Magic UKTV Gold MTV UKTV Gold +1 MTV Base UKTV Gold 2 MTV Dance UKTV Gold 2 +1 MTV Hits VH1 MTV2 VH1 Classic Music Choice VH2 Paramount

4 ANNEX 2 NON-CONFIDENTIAL VERSION

APPENDIX 9: CHANGES IN THE QUALITY OF SKY’S SPORTS AND FILM CHANNELS OVER TIME

A. Sky’s sports channels

9.1 Ofcom’s assessment of the changes in the quality of Sky’s sports channels since 2000 is as follows:

“although there has been some growth in the overall number of hours broadcast, the number of FAPL games shown by Sky has actually declined following the 2006 auction”. 85

9.2 As noted above, this analysis is too cursory for any evidential weight to be placed on it. Accordingly, we set out below evidence in relation to a range of improvements in the quality of Sky’s sports channels since 2000.86

9.3 In relation to FAPL football, as noted above, the assertion that the number of FAPL matches shown live on Sky’s sports channels has “actually declined” following the 2006 auction is erroneous. The number of matches broadcast on Sky’s sports channels has increased, from 88 to 92. Moreover, Table 9.1, below, shows that:

(i) the number of live FAPL matches broadcast on Sky’s sports channels has increased very materially since 2000 – from 60 matches per season in 2000/01 to 92 in 2007/08; and

(ii) the phrase “some growth” is an inadequate description of the growth in hours of FAPL football that is broadcast on Sky’s sports channels. Between 2000/01 and 2006/07 the number of hours broadcast increased by 256%.

85 Paragraph 4.27 of Annex 13. 86 We do not include the development of new high definition versions of Sky’s Sports channels as currently these require an additional subscription. ANNEX 2 NON-CONFIDENTIAL VERSION

Table 9.1: FAPL matches broadcast on Sky’s sports channels

Number of live Total hours of live FAPL FAPL matches programmes broadcast on broadcast on Sky Sky Sports channels Season Sports channels 2000-2001 60 246 2001-2002 66 449 2002-2003 66 465 2003-2004 66 589 2004-2005 88 682 2005-2006 88 624 2006-2007 88 636 2007-2008 92 n.a.

9.4 These data do not include matches shown via PPV, which comprised 40 live matches per season between 2001/02 - 2003/04, and 50 live matches per season between 2004/05 – 2006/07. (These matches were broadcast on Sky’s ‘PremPlus’ pay per view channels, not Sky’s pay TV channels.)

9.5 They also do not include matches shown on a ‘near-live’ basis. Broadcasting matches on a near live basis was introduced in 2004/05 and has resulted in an expansion in the number of both FAPL matches broadcast on Sky’s sports channels, and the number of hours of FAPL broadcast on those channels.

B. Other quality changes

(i) Number of channels

9.6 Sky Sports Xtra became a fourth Sky sports channel during 2001. Prior to that time it had been used to simulcast sporting events shown on Sky Sports 1, 2 and 3 with the addition of interactive services.

(ii) Additional programming

9.7 Between 2000 and 2007 there have been several additions to the Sky Sports programme offering including:

• live UEFA Champions League football, Sky’s coverage of which commenced in 2003.

• all live coverage of all international and domestic cricket played in England and Wales for the four seasons from 2006-2009, including exclusive coverage of all seven summer test matches (previously Sky Sports broadcast one summer test match and all one day home matches, as well as various overseas Test matches).

2 ANNEX 2 NON-CONFIDENTIAL VERSION

• all live coverage of Heineken Cup rugby in the UK and Ireland from 2006.

9.8 A summary of the changes to content on Sky Sports between 2000 and 2007 is provided in Table 9.2 below.

Table 9.2: Key changes to Sky Sports content 2000-2007

Sport Gains Not Renewed

Football Champions League (100 Scottish Premier League, additional games), Wales England Home Friendlies Home Internationals, Northern Ireland Football, South American World Cup Qualifiers

Cricket Seven England International Tests, Domestic Twenty/20 Cricket, India v. England Test Series

Golf US Tour Golf

Rugby Union European Cup Rugby (60 England Six Nations (3 matches) matches)

Rugby League National Rugby League Australian Rugby League

Motorsport A1 Grand Prix, Indy Car Series

Tennis Masters Series (10 events)

Darts Premier League Darts

Snooker Premier League Snooker

NFL 100+ additional games per NFL Europe season

Other Horse of the Year, Hickstead, British Basketball America’s Cup Sailing

(iii) On-screen improvements introduced since 2000

9.9 There have also been numerous improvements in the on-screen coverage of sports since 2000. Sky Sports has continued to refine and improve the innovations introduced to football coverage and introduced new features to coverage of others sports, including:

• ‘Sky Scope’ was introduced to cricket coverage in 2001. This feature measures the exact flight path of the cricket ball to accurately assess if it would have hit the stumps;

3 ANNEX 2 NON-CONFIDENTIAL VERSION

• ‘Ultra Motion’ was introduced for cricket in 2004 and uses ultra-slow motion cameras to provide instant replays not possible with conventional cameras;

• in 2006, ‘Hi Motion’ was introduced to cricket and rugby union coverage providing instant replays of up to 300 frames per second during live matches;

• ‘Virtual Eye’ graphics were introduced for the 2006 Ryder Cup to provide analysis of the course and play using a model of the course, live ball position data, tee and pin locations and green contours;

• in May 2007, Sky Sports introduced UK television’s first infrared camera technology on coverage of England cricket Tests and One Day Internationals against the West Indies. The ‘Hot Spot’ technology determines whether contact is made with the ball by identifying and measuring the heat generate by a collision and producing an image of the precise point of impact; and

• in 2007, rugby union coverage included ‘Zoom Box’ technology for the first time, providing video referees and viewers a closer insight into play by magnifying replay footage. Also, new camera technology that provides pictures from each corner flag, giving a view along each tryline was introduced. It was the first time in worldwide rugby union broadcasting that such angles were made available not just to viewers, but also video referees.

(iv) Other improvements to Sky’s sports channels

9.10 Other improvements to Sky’s sports channels introduced since 2000 include:

• ‘Football First’ was introduced for the 2004-05 FA Premier League season, an interactive programme that replays extended highlights of the day’s featured Premier League match and highlights from every Premier League match played on the day, enabling the viewer to select any number of matches to view;

• In 2003, multi-match choice options were introduced for football coverage on the Sky Sports interactive service. In 2006, a ‘quad split’ option was introduced for the first time for the UEFA Champions League, enabling viewers to select to watch a match from eight simultaneous live matches or switch between them;

• the Sky Anytime on PC video on demand service also features Sky Sports highlights including every FA Premier League games shown on Football First, selected UEFA Champions League matches, rugby, golf and cricket. Sky Sports highlights are available on Sky Anytime on TV for free to Sky Sports 1 and 2 customers.

4 ANNEX 2 NON-CONFIDENTIAL VERSION

C. Sky’s film channels

9.11 Ofcom’s assessment of the changes in the quality of Sky’s film channels since 2001 is as follows:

“There does not appear to have been a significant change in the quality of the Movie package products over the period. In particular, the number of films broadcast on Sky Movie channels has remained relatively stable since 2001.”87

9.12 Clearly, in relation to film channels, the number of films carried is only one aspect of quality. Accordingly, such an analysis is too cursory for any evidential weight to be placed on it. We set out below evidence in relation to a range of improvements in the quality of Sky’s film channels since 2001.88

(i) Changes to the mix of films

9.13 Over the last five years Sky Movies has gradually reduced the number of titles it is contractually required to take from the major studios. This means Sky Movies can select and broadcast the best output from the major studios without being contractually required to take lower quality titles or those that have not performed at the box office. This has enabled Sky Movies to purchase and schedule content from independent film producers where desirable, providing greater choice and higher overall quality to customers.

9.14 In 2000, 4% of movies on Sky Movies were blockbusters (i.e., they took over $100m at the US box office and/or over £10m at the UK box office and/or more than 1.5m video rentals in the UK.) By 2007 this has increased to 6%.

(ii) Changes to the timing of the pay TV window

9.15 The movie windows in which films can be shown on subscription channels have moved forward. This is an important quality enhancement as it means titles are provided to customers at an earlier date post-cinematic release than in 2001. In 2001 the pay TV window ran from 18 to 33 months after cinematic release; by 2007 had moved forward to run from 12 to 27 months after cinematic release.

(iii) Introduction of PIN access to Sky’s film channels

9.16 The introduction in January 2006 of PIN access on Sky Movies channels during the day means 12 certificate and 15 certificate films can be scheduled before 8pm. This means a wider choice and higher quality films can be shown during the day as there is no longer a restriction that only enables PG and U rated films to be shown before 8pm.

87 Paragraph 4.72 of Annex 13. 88 We do not include the development of new high definition film channels as currently these require an additional subscription.

5 ANNEX 2 NON-CONFIDENTIAL VERSION

(iv) Sky Anytime

9.17 The Sky Anytime on TV video on demand service launched in March 2007 and features movie content for customers with an enabled set top box and Sky Movies subscription. Titles are made available through a dedicated section of the EPG.

9.18 Sky Anytime on PC was initially launched in January 2006 as a web-based application and then relaunched in December 2007 as a download application. It is available to DTH customers with a Sky Movies subscription and features titles from Sky Movies, Sky Box Office Movies and World Cinema titles. It was the first legal movie download service in the UK. There is no limit on the number of films that can be downloaded and a large number of titles are accessible at no extra charge to Sky Movies subscribers. Subscribers to dual movies can access all titles for free.

(v) Introduction of genre-based channels

9.19 Sky has undertaken a number of reorganisations of it film channels in order to better appeal to customers and, in particular, to enable customers to find easily films that they want to watch.

9.20 The last of these reorganisations, which took place in April 2007 has meant the introduction of eight, genre-based standalone channels, together with two mixed-genre channels which show the same films as those broadcast on Sky’s two HD film channels and Sky Premiere, which is dedicated to the newest films to enter the pay TV window.89 Previously, while Sky had a similar number of SD channels, Sky Movies essentially comprised (a) Sky Premier, and a number of multiplexes of that channel, (b) Sky Moviemax, and a number of multiplexes of that channel, and (c) Sky Cinema (together with a single multiplex). Accordingly, while the number of SD channels has stayed relatively constant, the April 2007 reorganisation comprises a very major change to the presentation of Sky’s film channels, and – particularly in light of the fact that it required significant changes to Sky’s contracts with the Hollywood studios – was a change of considerable magnitude.

9.21 This change was made in order to respond to a clearly identified customer satisfaction issue that it was difficult for customers to find the type of movies they wanted to watch amongst the former Sky Movies 1 to 9 channels. This direct response to consumer feedback improved the quality of the Sky Movies offering by making the service better fit consumer demand, improving the clarity of the movies offered and helping the search for films.

89 The genre-based channels are Comedy, Sci-Fi and Horror, Action & Thriller, Modern Greats, Classics, Family, Drama and Indie.

6